-- David Swink, FCTA board member / 2016-12-13
The 2016 Presidential election was clearly won by Donald Trump, with 306 electoral votes, as required by the U.S. Constitution. The Founders designed this ingenious scheme to prevent the kind of mob rule witnessed during the French Revolution, and noted that the U.S. in fact is a Republic, not a Democracy -- the States are supreme. It is not our purpose here to bemoan the lack of American History and Civics in our public schools, but to merely explain the basics of the Electoral College and to show how far the national popular vote can theoretically contradict the electoral vote.
Each state is given a number of electoral votes equal to the sum of their two U.S. senators plus their alloted U.S. congressmen. That's it! So the smaller populated states like Wyoming (2+1) are not overwhelmed by massively populated states like California (2+53). Presidential candidates are thus encouraged to campaign in those intermediate-sized and politically-balanced states like Wisconsin (2+8) and Michigan (2+14).
To show the details of the 2016 Presidential election, we used U.S. Election Atlas as our source of popular voting at the state level in the table below. As widely reported, Donald Trump won 306EV-242EV over Hillary Clinton, while Clinton won the total popular vote by approximately 48%-46% in the four-way race with the Libertarian and Green Party candidates. So the actual election showed a 2% popular vote skew in the opposite direction from the electoral vote.
But what if we took the total state-by-state popular vote and hypothetically reallocated those votes in a two-way race to demonstrate how far this skew could go? The last columnar grouping does this very thing, by 1) giving Clinton the states Michigan and Pennsylvania to reduce Trump's win to 270EV; 2) postulates that all of Trump's wins were by a mere one or two popular votes; and 3) postulates that all of Clinton's wins were unanimous. (Maine is a special case, with a split EV, so Trump and Clinton are postulated to equally receive the third-party votes.)
And the result of this thought experiment? Donald Trump wins barely 270EV-268EV over Hillary Clinton, while Clinton wins the total popular vote by an astounding 75%-25% in the hypothetical two-way race.
Remember, this is merely an exercise in numbers. So while the 270EV minimum specified in step one is quite possible, the massive number of one-vote margins and unanimous state wins in steps two and three are clearly not possible. So the popular vote skew from the electoral count will always remain in the low single digits.